AI scams: new wave of deception in the digital age

In April 2023, Jennifer DeStefano from Arizona claimed a scammer used artificial intelligence (AI) to clone the voice of her daughter in a $1 million kidnapping scam.

Ms DeStefano told NBC she heard her "daughter's voice crying and sobbing saying ... mum, these bad men have me, help me, help me."

A man then got on the phone and said, "listen here, I've got your daughter. This is how it's going to go down.

"You call the police, you call anybody, I'm going to pop her so full of drugs. I'm going to have my way with her and I'm going to drop her off in Mexico."

According to the Australian Competition and Consumer Commission (ACCC) , financial loss from scams went up by 80 per cent last year.

And in the first quarter of 2023 (Jan-Mar), internet security company Norton blocked more than 25 million threats in Australia, equating to more than 277,000 threats a day, including 1.9 million phishing scams.

Norton's APAC director Mark Gorrie said with the evolution of AI video, voice and text models, picking a scam would become more difficult.

"There's obviously a lot of good here with AI and there is some really strong application for this, but whenever there's good, there's always a bad element," he said.

"Because scams are really social engineering techniques to trick people into giving up information, AI has allowed them to get a lot better."

Mr Gorrie said with the release of ChatGPT, it was easier for scammers to impersonate companies.

"Scammers can use the legitimate messages that have come from that particular company, feed it into a large language tool, and then create some different context and then embed if they're using malicious links in messages," he said.

"People thought, I know how to pick a scam message, they're poorly written, the grammar is bad, it has a lot of spelling mistakes.

"But with the use of these tools, a lot of that goes away.

"The messages look far more professional and obviously the output is very natural in terms of the text, so it does look far more legitimate."

IN OTHER NEWS:

Scams are also becoming more personal.

"With the scale of breaches we're seeing, there's so much data on the dark web.

"So there might be personalized content to the scam, whether it's using your name or an account number."

Mr Gorrie said more people will be scammed.

"In the past I think elderly people were much easier to trick," Mr Gorrie said.

"But I think as things improve and look more legitimate, it's going to catch more people out.

"I think people have to question why would I be getting that message?

"They're asking for particular information - It doesn't seem right.

"So if you are suspicious in any way, best follow it up, check it out, look up the details of whatever company or service is asking for that information and check yourself."